INFO-H-500 Image Acquisition & Processing
  • Home
  • 1.Introduction
    • Biological vision
    • Image sensors
    • Image representation
  • 2.Low-level image processing
    • Histogram
    • Linear filtering
    • Rank based filters
    • Image restoration
    • Edge detection
  • 3.Image segmentation
    • Typical image processing pipelines
    • Histogram based segmentation
    • Border based segmentation
    • Region based segmentation
    • Model based segmentation
      • Model based segmentation
      • Active contours
      • Hough transform
    • Examples
  • 4.Morphomathematics
    • Morphomathematical operators
    • Combined operations
    • The watershed transform
    • Gray level morphology
  • 5.Objects features
    • Statistical features
    • Contour features
    • Object moments
    • Texture features
  • LABS
  • References
  • About
  • Search
  • Previous
  • Next
  • GitHub
In [1]:
%matplotlib inline
from IPython.display import HTML,Image,SVG,YouTubeVideo
In [2]:
from skimage import data
import numpy as np
from skimage.morphology import disk
import skimage.filters.rank as skr
from skimage.measure import label
from skimage.morphology import watershed
from skimage.io import imread
from scipy import ndimage as ndi
import matplotlib.pyplot as plt
from skimage.segmentation import mark_boundaries
In [3]:
# segment the coins
im = data.coins()
plt.imshow(im,cmap=plt.cm.gray)
plt.colorbar();
In [4]:
# detect the eyes / nose
im = data.chelsea()
plt.imshow(im);
In [5]:
# counting the galaxies
im = data.hubble_deep_field()
plt.imshow(im);
In [6]:
im = data.page()

bg = skr.median(im, disk(10))

res = (1.*im/bg) < .8

plt.imshow(im,cmap=plt.cm.gray)
plt.colorbar();
plt.figure()
plt.imshow(bg,cmap=plt.cm.gray);
plt.colorbar()
plt.figure()
plt.imshow(res.astype(np.uint8),cmap=plt.cm.gray);
plt.colorbar();
In [7]:
# segment the cells
im = imread('../data/dh_phase.png')
th = im>150
th1 = im>100

plt.imshow(im,cmap=plt.cm.gray)
plt.colorbar();
plt.figure()
plt.imshow(1.*th+th1,cmap=plt.cm.gray)
plt.colorbar();
In [8]:
from skimage.feature import canny


ca = canny(im)

plt.figure(figsize=[10,10])
plt.imshow(ca,cmap=plt.cm.gray);
In [9]:
from skimage.morphology import watershed
from skimage.segmentation import mark_boundaries
lab,n_lab = label(th,return_num=True)
bg = th1==0
lab[bg] = n_lab+1

#med = skr.median(im,disk(5))
#gr = skr.gradient(med,disk(3))

ws = watershed(255-im,lab)
plt.imshow(mark_boundaries(im,ws))
/home/olivier/.conda/envs/py3/lib/python3.7/site-packages/skimage/morphology/_deprecated.py:5: skimage_deprecation: Function ``watershed`` is deprecated and will be removed in version 0.19. Use ``skimage.segmentation.watershed`` instead.
  def watershed(image, markers=None, connectivity=1, offset=None, mask=None,
Out[9]:
<matplotlib.image.AxesImage at 0x7f0e7a65bf90>
In [10]:
im = imread('../data/exp0001.jpg')
plt.figure(figsize=[20,20])
plt.imshow(im,cmap=plt.cm.gray)
plt.colorbar();
In [11]:
# count red and yellow flowers
im = imread('../data/flowers.jpg')
plt.imshow(im)
plt.colorbar();
In [12]:
# find the fiber orientation
im = imread('../data/image4.png')
plt.imshow(im,cmap=plt.cm.gray)
plt.colorbar();
In [13]:
from IPython.display import YouTubeVideo
YouTubeVideo('PUcz11MLxUk', start=0, autoplay=1, theme="light", color="blue",)
Out[13]:
In [14]:
# detect stroma
im = imread('../data/Rp042826d.jpg')
plt.imshow(im,cmap=plt.cm.gray)
plt.colorbar();
In [15]:
# segment the flowers
im = imread('../data/KaneFlowers.jpg')
plt.imshow(im,cmap=plt.cm.gray)
plt.colorbar();
In [16]:
from skimage.morphology import watershed
from skimage.segmentation import mark_boundaries

gr = skr.gradient(im,disk(3))

local_min = im <= skr.minimum(im,disk(5))

lab = label(local_min)

#med = skr.median(im,disk(5))

ws = watershed(gr,lab)

plt.figure(figsize=[10,10])
plt.imshow(mark_boundaries(im,ws))

#plt.imshow(local_min)
/home/olivier/.conda/envs/py3/lib/python3.7/site-packages/skimage/morphology/_deprecated.py:5: skimage_deprecation: Function ``watershed`` is deprecated and will be removed in version 0.19. Use ``skimage.segmentation.watershed`` instead.
  def watershed(image, markers=None, connectivity=1, offset=None, mask=None,
Out[16]:
<matplotlib.image.AxesImage at 0x7f0e73f73210>
In [17]:
rgb = imread('../data/4colors.JPG')

plt.figure(figsize=[20,20])
plt.imshow(rgb)
plt.colorbar();
In [18]:
r = skr.median(rgb[:,:,0],disk(1))
plt.imshow(r,cmap=plt.cm.gray)
Out[18]:
<matplotlib.image.AxesImage at 0x7f0e733b9e90>
In [19]:
s = rgb.sum(axis=2)
th = s > 100

#post-processing
pth = skr.minimum(th.astype(np.uint8),disk(1))

plt.figure(figsize=[20,20])
plt.imshow(pth,cmap=plt.cm.gray)
plt.colorbar()
Out[19]:
<matplotlib.colorbar.Colorbar at 0x7f0e7800cbd0>
In [20]:
lab = label(pth)

lut = np.arange(0,np.max(lab)+1)

plt.imshow(lab)
plt.colorbar()

mask = lab == 20
plt.imshow(mask)
Out[20]:
<matplotlib.image.AxesImage at 0x7f0e78a22e10>
In [21]:
from random import shuffle
shuffle(lut)
In [22]:
shuffle(lut)
plt.imshow(lut[lab])
plt.colorbar()
Out[22]:
<matplotlib.colorbar.Colorbar at 0x7f0e78cea290>
In [ ]:

In [23]:
# segment the cell
im = imread('../data/exp0001crop.jpg')
plt.imshow(im,cmap=plt.cm.gray)
plt.colorbar();
In [24]:
m = skr.median(im,disk(5))
plt.imshow(m,cmap=plt.cm.gray)
plt.colorbar()
Out[24]:
<matplotlib.colorbar.Colorbar at 0x7f0e789460d0>
In [25]:
th1 = m < 90

th2 = np.bitwise_and(110 > m,m < 130)

plt.imshow(th2)
Out[25]:
<matplotlib.image.AxesImage at 0x7f0e78a17810>
In [26]:
markers = label(th2)
plt.imshow(markers)
plt.colorbar()
Out[26]:
<matplotlib.colorbar.Colorbar at 0x7f0e78b0cbd0>
In [27]:
markers[markers==3] = 2
ws = watershed(im,markers)
/home/olivier/.conda/envs/py3/lib/python3.7/site-packages/skimage/morphology/_deprecated.py:5: skimage_deprecation: Function ``watershed`` is deprecated and will be removed in version 0.19. Use ``skimage.segmentation.watershed`` instead.
  def watershed(image, markers=None, connectivity=1, offset=None, mask=None,
In [28]:
plt.imshow(ws)
plt.imshow(mark_boundaries(im,ws))
Out[28]:
<matplotlib.image.AxesImage at 0x7f0e78c22610>
In [29]:
# segment the cell
im = imread('../data/brain.jpg')[:,:,0]
plt.figure(figsize=(10,10))
plt.imshow(im,cmap=plt.cm.gray)
plt.colorbar();
In [30]:
plt.hist(im.flatten(),255);
In [31]:
from skimage.filters import threshold_otsu

t_otsu = threshold_otsu(im)
t_otsu
Out[31]:
36
In [32]:
th = im > t_otsu
plt.figure(figsize=(10,10))
plt.imshow(th)
Out[32]:
<matplotlib.image.AxesImage at 0x7f0e78bcff10>
In [33]:
lab = label(th,connectivity=1)
plt.imshow(lab)
Out[33]:
<matplotlib.image.AxesImage at 0x7f0e730b1e10>
In [34]:
from skimage.measure import regionprops
In [35]:
props = regionprops(lab)

brain = (lab==7).astype(np.uint8)

pp = skr.maximum(brain,disk(3))
pp = skr.minimum(pp,disk(3))

plt.imshow(pp)
Out[35]:
<matplotlib.image.AxesImage at 0x7f0e7302c4d0>
In [36]:
for p in props:
    print(p.area, p.label)
1459 1
5 2
1 3
3 4
1 5
16 6
6323 7
1 8
2 9
1 10
1 11
1 12
1 13
16 14
1 15
1 16
1 17
2 18
2 19
2 20
2 21
30 22
1 23
1 24
1 25
1 26
2 27
2 28
5 29
1 30
1 31
2 32
1 33
13 34
In [ ]:

In [ ]:

In [ ]:


Creative Commons License maintained by Olivier Debeir and contributors.

Documentation built with MkDocs.

Search

From here you can search these documents. Enter your search terms below.

Keyboard Shortcuts

Keys Action
? Open this help
n Next page
p Previous page
s Search